Vascular dementia--still overdiagnosed.

THE NOTION THAT SLOWLY progressive dementia in the elderly is the result of chronic diffuse brain hypoxia is largely discredited — Alzheimer disease has nothing to do with cerebral arteriosclerosis.' Nonetheless, it is still widely believed that many elderly people are demented on a cerebrovascular basis, most often in the form of "multi-infarct dementia.' ' Such a term does not refer to the circumscribed intellectual disturbances which follow single large infarcts in the territories of major cerebral vessels. Rather it denotes more global cognitive loss in the presumed presence of several lesions, often deep lacunes, none of which would alone be expected to cause intellectual impairment. In fact, many clinicians believe that when a demented patient has symptoms or signs of stroke (as reflected, for example, in the Hachinski Ischemic Score) , cerebrovascular disease can be presumed to be either causing or contributing to the dementia, and on such an assumption rest a number of well-known clinical studies. Perhaps, therefore, one should ask upon what basis such an assumption is made. Tomlinson, Blessed, and Roth' have provided the most persuasive pathological evidence to date that stroke does cause dementia in the elderly. Comparing 50 brains of demented old people to brains of undemented controls, they concluded that half had what today is called senile dementia of Alzheimer type (SDAT), 17% had either definite or probable arteriosclerotic dementia, and 18% had a combination of the two diseases. The authors stressed a number of important details, however. First, of the 9 patients considered to have arteriosclerotic dementia, three had between 60 and 82 ml volume of infarction, and the other six had volumes of 101 to 412 ml; only one of the controls had more than 60 ml of infarction, and none had more than 100 ml. Second, the authors did not deal with "the possibility of ischaemic lesions affecting areas of brain likely to be particularly important in relation to producing features of dementia'' — possibly relevant to the fact that three patients with lesions of 60 to 82 ml had dementia, whereas the single control with such a degree of infarction did not. Third, four of the nine cases of mixed arteriosclerotic dementia and SDAT also had more than 100 ml of softening; the other five did have volumes of infarction (mean 38.8 ml) encountered in several controls, suggesting to the authors that the presence of either disorder might lower the threshold for clinical expression of the other. Moreover, softening of mild to moderate degree was present in 19 of the 25 patients considered to have